Just why do you insist on higher points to define a plan? I think that the CMM calculates the less square geometry (plan , line,cercle,..) doesn't it or Am I wrong?
to establish the datums in CMM we use mini max and Outer Tangential Element algorithms rather than LSQ to be sure the parts well fit together the LSQ algorithm provide a consistent stable result but consistently worng size , form and location .
He just explaine this in the video. If you only select 3 ... 2 ... 1 points to establish your datum system, there is no guarantee that you are going to use the points that will define the proper location, as defined by the standard (the associated feature should be the one that is the closest to the real feature, without going inside of the material). In fact, it is almost guaranteed that you will not. In that way, while you will of course have a working and stable datum system, it will not actualy reflect the functionality of the part in the real environment, so all the measurements you do with respect to that datum are deceiving and do not actually properly tell you if the part is OK or not.
There is a big difference between ISO GPS and GD&T: Association of a simulated planar touch is very different. This example works with a part which ca be orientated four times against specified datum system (it‘s two times symmetric). Under ISO GPS verification can look for the best result. The other orientations must not be ok.
I would argue that you need many more points than 6 to capture all the details of an integral feature. I am not sure in ASME, but in ISO, there is a standard (I can't remember the number out of the top of my head) that explains the extraction process. It is mentioned there how you should choose the point pattern, but also the number of points. While you cannot find an exact number (it would not make any sense), the requirement is that you should choose as many points as needed to capture all macro details of the surface, so that you would be able to accurately reconstruct the surface based on those points. In other words, someone should not be able to find a point outside the ones that you chose, that would change the location or orientation of the associated feature. I think this is something that is missed most of the time by measurement engineers from my experience. You could of course use datum targets to specify a desired 3-2-1 point datum system, but my oppinon is that you should only do that if that is the real case function of the part, or if you have a complicated free-form shape (like a turbine blade), that you need to inspect the surface of and you have no straight surfaces to use as datums.
Just why do you insist on higher points to define a plan? I think that the CMM calculates the less square geometry (plan , line,cercle,..) doesn't it or Am I wrong?
to establish the datums in CMM we use mini max and Outer Tangential Element algorithms rather than LSQ to be sure the parts well fit together the LSQ algorithm provide a consistent stable result but consistently worng size , form and location .
He just explaine this in the video. If you only select 3 ... 2 ... 1 points to establish your datum system, there is no guarantee that you are going to use the points that will define the proper location, as defined by the standard (the associated feature should be the one that is the closest to the real feature, without going inside of the material). In fact, it is almost guaranteed that you will not. In that way, while you will of course have a working and stable datum system, it will not actualy reflect the functionality of the part in the real environment, so all the measurements you do with respect to that datum are deceiving and do not actually properly tell you if the part is OK or not.
There is a big difference between ISO GPS and GD&T: Association of a simulated planar touch is very different.
This example works with a part which ca be orientated four times against specified datum system (it‘s two times symmetric). Under ISO GPS verification can look for the best result. The other orientations must not be ok.
I would argue that you need many more points than 6 to capture all the details of an integral feature. I am not sure in ASME, but in ISO, there is a standard (I can't remember the number out of the top of my head) that explains the extraction process. It is mentioned there how you should choose the point pattern, but also the number of points. While you cannot find an exact number (it would not make any sense), the requirement is that you should choose as many points as needed to capture all macro details of the surface, so that you would be able to accurately reconstruct the surface based on those points. In other words, someone should not be able to find a point outside the ones that you chose, that would change the location or orientation of the associated feature.
I think this is something that is missed most of the time by measurement engineers from my experience.
You could of course use datum targets to specify a desired 3-2-1 point datum system, but my oppinon is that you should only do that if that is the real case function of the part, or if you have a complicated free-form shape (like a turbine blade), that you need to inspect the surface of and you have no straight surfaces to use as datums.
Hello, can you share wiht us the number of this standard. Thank you.
@@marciosilveiramorais6983 Hi, I checed on my work computer this morning, it is "ISO 14406:2010 - Geometric product specification (GPS) - Extraction"